DESCRIPTION: Protein bioinformatics provides the mathematical and statistical tools necessary to analyze proteomic data. It is therefore a key field for progress in health-related biomedical research. The work proposed herein is designed to develop methods for minimizing the loss of information in bioinformatic computations. The specific thrust of this work is the development of new approaches to sequence alignment and homology searching.